On-farm data-rich analysis explains yield and quantifies yield gaps of winter wheat in the U.S. central Great Plains

Abstract

With an annual production of ∼60 Mt, the U.S. accounts for about 8% of the global wheat (Triticum aestivum L.) production. Still, quantification of the yield gaps (YG) and major management factors to reduce it are scarce. We used Kansas, the largest wheat producing state in the U.S. located in the central Great Plains, for an initial assessment of on-farm yield and YG. We collected field-level management (37 variables), weather (8 variables), soil (two variables) and yield data from 656 commercial wheat fields over three harvest years (2016–2018) to (i) quantify management adoption levels, Ya, and YG, and (ii) identify interactions among management practices and weather variables using a data-rich approach. We also used our data as a case-study to detect whether differences in crop management among regions justified data clustering by crop zones. Water-limited yield potential (Yw) was simulated for each field-year using actual soil and weather data and the SSM-Wheat model. Fields were grouped in three climate zones based on their long-term climatology and important differences in cropping systems between zones. Grain yield averaged 3.8 Mg ha−1 and ranged from 0.3–7.1 Mg ha−1 across all regions and years. The YG averaged 44 %, with seasons with high Yw resulting in greater YG. Management practices most often associated with grain yield were management of nitrogen (N), phosphorus (P), and sulphur (S) fertilizer, as well as foliar fungicide and its interaction with variety reaction to major diseases, although these depended on in-season weather. Our analyses highlighted many other genotype × management × environment interactions explaining winter wheat Ya, such as regional-specific cultivar maturity and the dependency of sowing date (and its relation to seeding rate) on cropping system.